Skip to content

Benchmark

Arena ELO - Math

Math benchmarks measure how well models solve maths problems. Higher scores are better.

Models measured
350
Scores recorded
405
Sources
1
IQ weight
1.24

Score distribution

1,316
Mean
1,367.4
Median
152.6
Std dev
1,089
10th percentile
1,474.8
90th percentile
837.8
Min
837.81,547.6
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.24 for chat models, measured over 350 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
51MiMo V2.6 Flash· LMArena data1,468.1
52Claude Sonnet 5High· LMArena data1,466.4
53Qwen3.7 Plus· LMArena data1,466.1
54Step 5 Preview· LMArena data1,465.6
55Claude Opus 4.5High, 32K, 20251101· LMArena data1,463.4
56Qwen3.6 Max Preview· LMArena data1,461.5
57Claude Sonnet 4.6· LMArena data1,461.5
58GPT-5.6 TerraxHigh· LMArena data1,461.4
59Gemma 4 31B IT· LMArena data1,461.1
60Beam 1.0 501B A23B· LMArena data1,460.9
61Qwen 3.8 27B· LMArena data1,458.9
62Inkling Small· LMArena data1,458.7
63Claude Opus 4.520251101· LMArena data1,458.4
64Kimi K3.1 ProMax· LMArena data1,458.4
65Grok 4.5· LMArena data1,457.4
66Muse Spark· LMArena data1,457.3
67GPT-5.6 LunaxHigh· LMArena data1,457
68GPT 5.4· LMArena data1,456.7
69Nemotron 3 UltraNVFP4· LMArena data1,456
70GPT-6 AstraMax· LMArena data1,455.6
71Grok 4.20Reasoning, 0309, beta· LMArena data1,454.2
72DeepSeek-V4-Pro PreviewHigh· LMArena data1,453.9
73Qwen3.5 397B A17B· LMArena data1,453.8
74Gemini 2.5 Pro· LMArena data1,450.4
75GPT-6.1 TerraMax· LMArena data1,450.4
76Claude Sonnet 4.5High, 32K, 20250929· LMArena data1,449.1
77Qwen3.6 Plus· LMArena data1,447.5
78Gemini 3 FlashMinimal· LMArena data1,447.1
79MiMo V2 Pro· LMArena data1,447.1
80GPT 5.1High· LMArena data1,446.8
81Qwen3 Max Preview· LMArena data1,446.8
82Mistral Large 4· LMArena data1,444.9
83Kimi K2.5Instant· LMArena data1,442.8
84DeepSeek V4 Pro· LMArena data1,441.4
85LongCat Flash Chat· LMArena data1,440.7
86Grok 4.20 Multi-Agent0309, beta· LMArena data1,440.7
87Qwen3 Next 80B A3B Instruct· LMArena data1,440
88GPT 5.2High· LMArena data1,439.9
89GLM 5· LMArena data1,439.3
90nova-experimental-chat26-02-10· LMArena data1,439.3
91Doubao Seed 2.0 Pro· LMArena data1,438.5
92Qwen3 Max2025-09-23· LMArena data1,436.2
93ERNIE 5.00110· LMArena data1,435.9
94MiMo V2.5· LMArena data1,435.3
95DeepSeek V3.2· LMArena data1,434.8
96Gemini 3.5 Flash Lite· LMArena data1,434
97GPT-5.2 Chat20260210· LMArena data1,433.7
98Grok 4.20 Beta 1· LMArena data1,433
99Claude Haiku 5.5High· LMArena data1,432.9
100LongCat-Flash-Chat Exp2602· LMArena data1,432.2

Showing 51–100 of 405 results.

Data credits: LMArena, CC BY 4.0 (licence)